Under the Third Generation Partnership Project (3GPP) release 99 framework, a radio network controller (RNC) may control resources and user mobility. Resource control may include admission control, congestion control, and channel type switching. Uplink data may be allocated to an Enhanced Dedicated Channel (E-DCH), which may include an Enhanced Dedicated Physical Control Channel (E-DPCCH) for data control and an Enhanced Dedicated Physical Data Channel (E-DPDCH) for data. The E-DPCCH and the E-DPDCH may be discontinuous and may be transmitted only when there is uplink data to be sent. Additionally, uplink data may be transmitted on a continuous Dedicated Physical Data Channel (DPDCH). A radio base station (RBS) may include an uplink scheduler that determines which transport formats each subscriber may use over the E-DPDCH.
As previously mentioned, the RNC may be responsible for admission control and congestion control. For example, the RNC may monitor and control the load in the RBS. The RNC may perform these operations based on Iub interface measurements from the RBS. The Iub measurements related to the uplink may include received total wideband power (RTWP) (i.e., the total received power at the uplink receiver), reference received total wideband power (RRTWP) (i.e., the thermal noise contribution to the RTWP), and received scheduled E-DCH power share (RSEPS) (i.e., the received power from resources controlled by an enhanced uplink (EUL) scheduler (e.g., the E-DPCCH and the E-DPDCH) relative to the RTWP). In one implementation, the Iub measurements may be transmitted to the RNC by the RBS in a Node B Application Part (NBAP) report. In some instances, the NBAP report may include both the RSEPS and the RTWP for the same time interval to enable direct comparisons.
FIG. 1 illustrates an exemplary uplink stack 100 that includes exemplary uplink interference contributions. As illustrated in FIG. 1, total uplink interference (I-total) 140 may include background noise interference 105, other-cell interference 110, DPDCH interference 115, DPCCH interference 120, non-scheduled interference 125 and scheduled interference 130. Non-scheduled interference 125 may include interference from the E-DPCCH, the E-DPDCH, and a High Speed Dedicated Physical Control Channel (HS-DPCCH). The HS-DPCCH may be employed for uplink acknowledgements relating to downlink data transmitted over a High Speed Downlink Shared Channel (HS-DSCH). Scheduled interference 130 may include interference from the E-DPCCH and the E-DPDCH. The interference contribution of scheduled interference is further illustrated by uplink scheduled interference (I_sch) 135.
Based on measurements over the Iub interface, the following may be estimated according to the following expressions:Uplink noise rise as =RTWP/RRTWP; Uplink relative load as Lnr=1−(1/)=1−(RRTWP/RTWP); andNon-scheduled load as Lnon-sched=Lnr−RSEPS.
In such an instance, the non-scheduled load estimate may include the load due to inter-cell interference from other cells. Additionally, E-DCH may yield a non-scheduled load because the DPCCH of the E-DCH may be considered non-scheduled.
When balancing scheduled and non-scheduled loads, the non-scheduled load may be used as input to the admission control of the RNC to ensure that there is sufficient headroom for scheduled data. This reallocatable resource intended for scheduled E-DCH is referred to as the scheduling headroom. This may be expressed as:Lsched, headroom=Lnr, max−Lnon-sched,where Lnr, max is the maximum uplink relative load of the cell based on, for example, a coverage or power control stability metric.
For a target scheduling headroom, a target non-scheduled load, Lnon-sched, target of the cell may be derived, to which an estimated current, non-scheduled load may be compared. In such a comparison, an admitted load, Ladm, from recently admitted connections that are still inactive may be included. Consequently, a user may be admitted if the following expression is met:Lnon-sched+Ladm+Lnew potential connection≦Lnon-sched, target.
Margins considered by, for example, a load estimation algorithm (LEA) and/or a scheduler may affect the available scheduling headroom. For example, the RNC may employ a LEA for purposes of admission and/or congestion control. Additionally, or alternatively, the RBS may employ a LEA for scheduling, and/or assign grants to subscribers based on the scheduler. The LEA may calculate the load contribution from non-scheduled connections in their own cell, Lnon-sched, own, and may maintain an estimate of other-cell load contribution, Lother (i.e., the other-cell received power share). For example, the other-cell load contribution may equate to a ratio between received powers from other cells and the RTWP. In this regard, the scheduler may consider the scheduling headroom according to the following expression:Lsched, headroom=Lnr, max−Lnon-sched, own−Lother.
Further, in order to maintain a margin for inter-cell interference, and to be robust to estimation errors of the other-cell load contribution, the other-cell load contribution may be limited from below by a minimum other-cell load contribution Lother min. In one implementation, Lother min may be a static value. Thus, the scheduler may consider the scheduling headroom according to the following expression:Lsched, headroom=Lnr, max−Lnon-sched, own−max (Lother, Lother min).
Such a margin, which is not always active, may not be accounted for in the Iub measurements. Additionally, there may be other margins utilized by the LEA and/or the scheduler that may not be accounted for in the Iub measurements, but reduce the scheduling headroom considered by the RBS. Consequently, such margins may not be known by the RNC and correspondingly may not be taken into account.
Additionally, multi-user detector schemes and/or interference cancellation schemes may be adopted by the RBS to cancel intra-cell interference. One approach to such schemes includes regenerating the interfering signal from detected connections and subtracting the regenerated interfering signal from the received signal. Thus, the effective interference power from an E-DCH may be less than the actual received power. Therefore, the RSEPS may not reflect the actual balance between the E-DCH and a DCH.